Traditional database methods utilize a direct mapping between a database schema and data structures stored in the database. In such methods, a given data structure may be stored as a single row in a database table, and the columns of the database table will need to match the data types within the data structure.
For example, a database could be configured to store a data structure that includes a string data type and an integer data type. The database would include a table with each instance of the data structure stored as a row in the table. Accordingly, the table would have to have at least one column configured to store strings, and at least one column configured to store integers.
A problem with this method is that a change to the data structure would require a change to the database table. For example, if an upgrade to the application changes the data types of fields in the data structure, then the database will need to be updated and will no longer be compatible with older versions of the data structure prior to the update.
Accordingly, database technologies may benefit from improved techniques of storing data structures in a database.